IBM MQ EUVD-2026-78151

Exploitation requires (1) a low-privileged, authenticated position in the IBM MQ environment - the CVSS vector specifies PR:L, so this is NOT an unauthenticated, default-configuration attack - and (2) the ability to supply XML content to an MQ component that parses it with external entity resolution enabled. The specific exploitable surface is whichever XML-consuming pathway the environment exposes (message payload parsing, JMS/XMS client input, MQ Managed File Transfer, or administrative XML/REST handling); defenders should confirm which of these are enabled and reachable in their deployment. Limits on exploitability: the attacker must first obtain valid MQ credentials or channel access, which confines realistic risk to insider or post-compromise scenarios; environments that disable DTD/external-entity processing in their parsers are not affected through that pathway; and any deployment that does not process untrusted XML at all has no practical exposure. Note that IBM does not describe a non-default setting that is required to be vulnerable - the flaw is in the parser behavior itself - so the practical gating factor is authenticated access plus an XML ingestion path.

Signals are mixed but lean toward a medium-to-high operational priority for MQ operators. The CVSS 3.1 vector (AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:N/A:L) rates confidentiality impact High and availability impact Low, with network reachability and low attack complexity - but crucially PR:L, meaning the attacker already needs a low-privileged, authenticated foothold in the MQ environment. That materially lowers the population of exploitable scenarios compared with an unauthenticated XXE. No EPSS score was provided in the input, no CISA KEV listing exists, and no public proof-of-concept was identified, so there is no evidence of widespread or opportunistic exploitation; treat this as a targeted-risk item rather than an emergency. Note two explicit data gaps: EPSS is missing, so exploit-probability cannot be quantified, and the absence of a concrete fix version means patch posture cannot be verified from the supplied references. The score is credible and not obviously inflated - an XXE that reads files is a genuine High-confidentiality finding - but it is not a 'drop everything' critical.

An attacker with a low-privileged, authenticated MQ connection (or access to an MQ subsystem that parses attacker-supplied XML, such as a message-producing client or an administrative XML interface) submits a document containing a malicious external entity declaration. The vulnerable parser resolves the entity, embedding the contents of a local file - or returning them to attacker-controlled output - and repeated or recursive entity expansion can additionally exhaust memory and degrade the queue manager. Since the vector is network-reachable with low complexity and no user interaction, the attack can be scripted, though no public exploit code was identified at time of analysis.